Autonomous Reality Capture for Industrial Facilities

- Good morning, everyone. Thank you for joining us for today's webinar. We'll be talking today about how you can automate laser scanning with SPOT and the Leica BLK ARC. Some quick housekeeping, before we get started, we'll be sending out a recording of today's presentation following the webinar. And if you have any questions during the webinar, please use the Q&A button. We'll be going through some of those towards the end of the presentation. With that, I'll hand it over to Johannes to introduce himself and kick us off. - Hello, everyone, and also a warm welcome from Iceland to today's webinar. I'm really overwhelmed by this large group joining us online. My name is Johannes Hotz. And I'm the Business Lead for the BLK ARC. In this role I manage our customers-partner relationships. And I perform hundreds of proof of concepts with robotics and laser scanning throughout the years. I'm excited to collaborate with you and work with Boston in future. A bit about my background, I'm a surveyor by trade. So I studied geodesy in Karlsruhe many, many years ago and worked now, nearly 25 years with Leica Geosystems. And I have spent my entire professional career in this geospatial industry. So that's me. - Julita, would you like to go next? - Yeah. Hello, everybody. My name is Julita. I work as a Product Engineer in the BLK product management team, where my main focus is on the BLK ARC. As of my background, I'm also a surveyor with many years of experience in the reality capture industry. - And my name is Brian Ringley. I'm a Principal Product Manager at Boston Dynamics, where I lead digital twin product strategy. And I've worked over the years in construction and architecture before joining Boston Dynamics on everything from design models to validating those models with as-built exercises that leverage the technologies we'll be discussing today. Now, we have been working with Leica and the BLK ARC scanner for a while. And we've actually found value across a pretty large number of industries. In fact, I'd argue that every industry we served has found some kind of value from our partnership and from autonomous laser scanning. So why are we gathered here today to have a webinar? Well, what we want to talk specifically about and kind of hammer home, I think, are three main points. One is the suitability of the Leica BLK ARC itself for mobile robotic transport. I think it is the most productized mobile laser scanner product out there. And it has been most specifically developed for the unique forces that are encountered by mobile carriers like SPOT. Two is the fact that this is the first time we've integrated laser scanning completely into our entire industrial inspection system. So this spans the robot integration itself, as well as our Autowalk tablet software and our Orbit Fleet Management and factory intelligence software. And then lastly, it allows us, I think, to really start to speak in earnest about the digital twin. I think most people here with a background in these technologies understands that laser scanning itself does not constitute a digital twin, but nonetheless is a critical part of building those twins. And I think what we're excited about today is that through this integration with Orbit, we're actually combining our industrial inspection package, which gives you your valuable asset sensor data streams with three-dimensional context of your facilities. And this, we think, starts a new generation of our industrial inspection offering, giving you a full picture of your factory. So what is this used for? What is a digital twin for? Well, it's for a little bit of everything. But I like to focus the conversation on how you get your foot in the door to start planning for this. Because this, like many other emerging technologies, is part of a digitization journey for your companies. So by and large, what we're focused on initially is facility documentation and automating that process. A lot of you have facilities that either predate modern BIM, or as anyone who's familiar with BIM knows, your BIM is out of date, or really any kind of model you've generated for construction onto operations of your facility gets out of date pretty quickly. That's because these are dynamic industries. You're either doing construction on your facility. Or the manufacturing lines themselves are changing based on the evolution of your product. And when you have documentation that is actually up to date, it opens up a bunch of new doors for value in your companies. And this is really what twin strategies are all about. The ability to access site information without the expense of travel or without the need to send personnel there physically is hugely valuable. I think the world really woke up to this by necessity during COVID, but these technologies have been advancing pretty steadily ever since. And then there's the asset management aspect. You only know as much as you ask. And it's one thing to be getting the streaming data of the temperature of an asset, the pressure reading, whatever the case may be. But often, to really answer serious questions and create factory intelligence, you need more context. You need the geometry of the asset itself. How does it look right now, where is it positioned, what else is happening in that space? And reality capture lends itself really well to that. And then ultimately, all of these things are used to better plan for change. And just at a baseline, having current information available always at your fingertips, without an extraordinary amount of expense or effort, allows you to better plan for that change. And it also sets the stage for more mature twin processes, where you can actually leverage model information to simulate operations to better make future decisions. So I think there's pretty broad consensus across industries that digital twin processes, or at least some related subset of these technologies and workflows, are valuable to a business. And this is something you need to be planning for today. And yet, when people are surveyed about the actual implementation of these strategies, the number of people who have done the thing, versus said the thing is valuable, is much lower. And many, many of those people in their businesses, that's a partial implementation, that's a POC, that's a pilot. It's not part of their day to day operations yet. And why is that? Well, I think very simply, it's because it's really expensive to do. And if something is super time-consuming and labor-intensive and that expense is rising, and you're going off of a future value hypothesis, then you're going to be a little bit more reluctant to make that investment. So at the very least, automation has value that is saving time, that's saving labor. I mean, that's something I think we all agree on is that we're entering an era of reality capture, where that capture can be automated and not labor-dependent. And in particular with something like laser scanning, that labor dependency can really cause a lot of friction, because laser scanning typically requires some kind of best practice. Even with mobile laser scanning, which is fairly plug and play and you're walking around, there are best practices to how you move that sensor through your space in order to capture a good data set. And the sophistication of the laser scanning sensors themselves don't really lend themselves very well to a fixed IoT system. You're probably not going to install your factory with 1,000 BLK ARC scanners, although I'm sure Johannes would love that. So the practicality of dynamic sensing with something like SPOT, in general, is the ability to move some of these more sophisticated sensors, where you don't need data at a millisecond or a Hertz rate, but you need data at, say, from a daily to a weekly or monthly rate. It really makes a lot more sense financially. And the back end matters too. So there's a lot of processing there. And I'll talk a little bit more about that later. So again, breaking this down, what's the value proposition? Well it's just an easier and cheaper way to get the same data sets that you're already getting. And if you're not already getting them because of the labor and cost requirements, this allows you to get into the game. The thing I'm most excited about is when you make laser scanning cheap to perform, it opens up the door to do it more frequently and to use it as a means of tracking change detection and maintaining a more up-to-date picture of your factory environment. And when you can do that without a corresponding increase in cost because we have this system that basically just lets you rinse and repeat these autonomous missions, that opens up all sorts of new use cases for laser scanning that just simply weren't feasible before. And we really see this again, as moving us into the next generation of industrial inspection features with SPOT, Autowalk, and Orbit. With this system, overall, it's a very seamless extension of what you're already doing. If you already have an installation set up for industrial inspection today and you have a site map today, it takes probably less than an hour to get a laser scanning mission out there because you can just utilize that existing information in Orbit, to basically offline create new scanning missions. And that's something that customers specifically asked for when we said, when we bring scanner to the table, what do you want? Well, we want everything controlled through one interface. And I also already have my site map set up. I've already installed my robot, so I want to leverage my existing installation to get this additional type of inspection for this full picture of my factory without more investment. I just want this new feature to work. And I think we've built something like that. And then lastly, for those of you who are already laser scanning, you're probably already using these Leica products. If you want to get value out of this data, as much as I love enabling getting this data as frequently and as much as possible, that's not inherently valuable. You have to analyze the data. You have to be answering business questions with the data, generating reports, generating insight that adds value to the business. This is also why we think it's so important to partner with Leica, who have a whole suite of software applications. And some of you may and probably are already using and can continue to use with our system. And if not, I think you're going to be excited when the Leica presenters go over that. So to recap when we're talking about the industrial inspection offering we have, we have a lot of customers today that have this installed. We have these four primary sensing modes. They're going around, they're looking at your assets, they're tracking these measurements against acceptable operational thresholds and alerting against those. But again, you're getting some pictures of a thing. And you're getting information and alerts about a thing. That is like one start to value. And now we're adding this more three dimensional context. If the robot is already walking around your factory and it's already taking all of these measurements with its sensors, why don't we just essentially, exhaust this reality capture data for you, and then marry the facility and the sensor data to start a twin strategy here? And that's again, I think part of our overall, digital journey here with SPOT and Orbit. All of this information is in one control interface. You don't have to learn new software to do this and to get up and running. It all runs through the same data pipes. So whatever is currently transmitting your asset information into Orbit is also taking those point clouds into Orbit. And then when you have all of that information consolidated in one place, it makes it really easy to retrieve that data and download it, or to use webhooks and integration to send it downstream to software like ERP systems and EM systems. So let's break down the actual workflow because I think people like the specifics here. We want a whole product solution, not a point solution. So yes, we've automated scanning. But no, that's not the whole story. And that does not create value in and of itself. So we're really looking at every step in the value chain here and making sure we're reducing labor across the board. The reason there's friction and expense for laser scanning is because there's labor at several steps, not just at the point of capture. So we start with route planning in Orbit. Again, if you already have your sitemap set up, you have routes you like to use, and you just want to use those same inspection routes to direct scanning missions, you can actually do that. And this is a really incredible feature, the ability to select any waypoint in your sitemap and convert that into a scanning location or a pass through location for a continuous scan. And I'll talk about the difference between those two things. If you don't, if scanning is actually your way to get into SPOT in Orbit, and that's why you're tuning in, that's awesome too. And that works the same way as setting up your site for inspections, where you take our tablet application out. You record the space with the robot. You can either add critical scanning locations or pass through waypoints in the field with the tablet. If you don't feel like doing that and you simply want to map navigation routes, then you can do all that other work offline in Orbit. So it's a super flexible system. And you'll see a couple images there to the right of our site map. And this is an example mission I set up. And I just want to point out the first piece there is continuous scanning. This makes sense. A mobile scanner on a moving robot means that robot is always in motion. It's maximizing coverage and facility data, relative to the time you have in your factory. And in this example, I had a critical asset, kind of in the middle there, where I had those white waypoints. So the system is flexible. I can continuously scan with the BLK ARC. And I can control resolution with that by controlling the speed at which the robot walks, also editable in Orbit, the same way I would walk around with a scanner and control my own speed. But I can also insert what are called scan static scanning locations using the same system. So in a case where you need more detail, either because it's an asset that's critical, maybe there's been recent change in a location, so you need more density there. Maybe you're prepping for some kind of construction or line change activity. This allows you to go into Orbit, make those changes very quickly to your existing scanning mission or missions, and then schedule those changes to run. So we pick up that data transfer as well, like I mentioned, that's available for you in Orbit. And then this is where the Leica software really shines. You're able to quickly download those data sets because they're in the same location where you're keeping your inspection data in Orbit. And then you're able to use those downstream, either in Leica's cloud products that are powered by HXDR, such as Reality Cloud Studio or in their Cyclone desktop apps. And again, Johannes and Julita are going to talk more about what can be done there. - Thank you Brian. So now let's hear a little bit more about the BLK ARC. So the BLK ARC, it's a 3D laser scanner with full 360 degree scanning capabilities. It allows us to capture detailed and highly accurate point clouds and to represent the scanned environment. And the device takes advantage of the Grand SLAM technology, which is a combination of dual axis LiDAR, multi-camera vision system, and IMU, which all combined allows to scan while we are in motion. And with the BLK ARC, we can also capture in dynamic and static modes, as Brian mentioned before. And the dynamic scanning allows us to capture the environment in a fast and efficient way, decreasing the capturing time by over 50%, in comparison to static scanners. This means that it's not only getting more done in less time for the operator, but also it's shorter interruptions on the side. And when it's all about capturing higher detail, the BLK ARC can also perform static scans, which helps to densify the point clouds. And the BLK ARC can capture up to 420,000 points per second. And thanks to the panoramic cameras, which are capturing images during the scan, we achieve fully-colorized point clouds. The lightweight BLK ARC, historically, it was designed for autonomous and robotic applications. It can be operated remotely with minimal user intervention required. It can be set up to scan a space without constant supervision, which again, it can significantly increase the productivity on site. And the scanner weighs below 700 grams, which makes it a perfect lightweight payload for the SPOT, not influencing the battery life of SPOT and leaving, of course, space to mount other payloads along. Now a little insight about the Grand SLAM technology. So SLAM stands for simultaneous localization and mapping. And our Grand SLAM solution is a fusion of LiDAR SLAM, visual SLAM, and the inertial measurement unit, where LiDAR SLAM uses geometric features, such as planes and the intersection of planes captured by the LiDAR, to calculate the BLK ARC's position. And the visual SLAM uses features, visual features, which are tracked and detected based on the images captured with the panoramic cameras during the scan. And all that combined with the IMU, it allows BLK ARC to self-orient itself, while capturing the reality. Now, after collecting the data, it can be very easily offloaded from the scanner, either by connecting directly to it or through Orbit or the SPOT app as well. And then afterwards, we provide two solutions which can be used for post-processing and registration of the data. The first being, our Reality Cloud Studio, which is cloud-based solution. And the second, our flagship software, the REGISTER 360, which is available locally on the workstations. Then once done, the data can be further managed in a variety of different software packages that can be provided by Leica. But also, other third party solutions can be used. And that means we can continue working in Reality Cloud Studio to create, for example, 3D meshes or virtual walkthroughs. And the data can be also imported into Cyclone 3DR, where for the point-cloud filtering can be applied, and we can model many, many elements. Or as I mentioned, the data can be also imported into third party software to produce other endless types of deliverables. Now, diving deeper into what is available in our flagship software. So starting from the REGISTER 360, which is a huge toolbox for post-processing the BLK ARC data, it allows us to visualize the data and register multiple scans together and while simultaneously tracking the quality of the registration so we have full control over them. And it also offers offline SLAM recalculation to take advantage of functions such as re-observation, to make sure that we achieve best possible results in terms of accuracy. Especially it matters in more challenging environments. And it also gives us the possibility to export professional registration reports, which can then be used as part of a project documentation. And of course, we can publish the post-processed scans in many file formats, such as BTS, LES, E57, RCP, and more. Then moving on to the Reality Cloud Studio, which is the cloud-based tool, allowing us again, to visualize the data, create virtual tools, annotation, measurements, all that shown in the 3D space. It enables direct data-based cooperation between office and field crews. It connects the service provider with their contractors, enables collaboration between teams and divisions within the company. And the Reality Cloud Studio also provides easy-to-use auto mesh and auto registration workflows. We can also export, of course, the originally uploaded file, the B2G file. And we can convert the point cloud into E57 LES, or BTS file formats as well, and of course, convert the point cloud into a 3D mesh model. Now, the cherry on top, Cyclone 3DR, which is filled with many, many functionalities, starting from cleaning and filtering point clouds, as well as auto-classification of objects within a point cloud. And this can be done automatically by used machine learning models. And those are available, both for outdoor and indoor scenarios, for example, to reduce noise within the point cloud or erase moving object. Those cleaning and filtering, it can also be done in manual way. So then the user decides what part of the data stays in and what has to be removed. And in Cyclone 3DR, we can also create 3D mesh models, perform as-built and clash analysis. We can model and create elements like digital terrain models, floor plans, sections, ortho images, topographic plans, pipe models. It's really endless. And just a few samples of deliverables that can be produced based on the data collected with BLK ARC. Starting from the beginning, so the point clouds, which can be used for quick analysis of the environment, for flash analysis, and many more, we can produce the virtual walkthroughs, 2D and 3D CAD models for various use cases, and as well as BIM models. And now I will hand over to Johannes, who will tell you more about our use cases and show you a few data samples. - Thanks, Julita. So I think the speaker before me really explained very well, the advantages of laser scanning, of robotics, of autonomy. I want to really go a little bit more into details and give some examples to give an idea of what is possible. Laser scanning enables fast and safe capture of every detail in environments, like manufacturing facilities, tunnels, buildings, or power plants. And the resulting point, cloud data offers invaluable insights for designers, engineers, operations, maintenance teams, and even offsite vendors, ensuring a perfect fit for a new tooling installation, for example. By incorporating laser scanning and reality capture, your digital twin transforms into a comprehensive information hub that enhances all aspects of your manufacturing operation, helping you reduce costs and increase efficiency. As mentioned earlier with the Boston Dynamics SPOT robot, laser scanning becomes autonomous, seamlessly integrating into your daily inspection routines. Laser scanning forms the foundation for various use cases. One key application is as-built documentation. For example, in industrial facilities, different user groups utilize 3D laser scans for their daily operations. A straightforward use case is in manufacturing operations. Imagine you needing to relocate equipment within your facility. How can you be sure it will fit through the doors and corridors? With a 3D scan of your facility, you can verify these details prior to any disassembly or ordering new equipment. Another example, as Brian already mentioned earlier is factory planners. 3D scans are a huge help, eliminating the need to travel to multiple sites. In addition to 3D scans, the BLK ARC can also capture panoramic images, providing a virtual walkthrough through the space. Health and safety use cases are another benefit. For instance, you can monitor escape paths daily and use point cloud to check if they are clear or obstructed, I think a very important element in today's world. And lastly, for facility management, the ability to accurately locate assets with your facility is crucial. Whether for renovation or maintenance, this capability really significantly improves the efficiency. Next. And the use of point clouds-- and here you see an example of a manufacturing facility. The use point clouds in capturing and manage the as-built status of facilities represents a significant step forward in facility management and planning. For example, as-built documentation, it provides a detailed representation of the facility's current state. Useful for maintenance, as I said before, operational management, reconstruction or renovations. Element classification and extraction, as Julita said before, predefined elements, AI-trained elements like doors, windows, floors, vegetation, and pipes can be identified and managed directly from the point cloud. And this supports inventory management by determining counts and dimensions, for example. And facility and production planning, another key element in this use case. It enables replanning of production sites with accurate spatial awareness. It assists in verifying equipment fitment, ensuring smooth integration of new machinery. Detailed visualization, it can capture complex details like ceiling installations, ventilation systems, pipes, cables, machinery, and even water hoses, as you have just seen in the video. Therefore, facilities will better understand and be able to analyze the local situation without the need of physical access to the location. And lastly, modeling and measurement. The captured point cloud can serve as a base for the 3D model, which can be created with the Leica software. And the technology enabler of this is the BLK ARC, combined with the Boston Dynamics SPOT and Orbit. This system provides high resolution point cloud captured, ensuring comprehensive visibility of structural and mechanical components and fully autonomous. I think that's a huge, huge thing that we can offer to our customer base, that these things become part of your daily or weekly routines. The key benefits for this one is I think, obvious, the efficiency gains. It reduces time and effort for individual inspection measurements on site. And Julita mentioned before, the interruption of the site, if you want to enter a RoboCell on an automotive plant, the RoboCell has to stop. I mean, no one can enter. So these interruption times are reduced, and therefore, the bigger efficiency gain. Accuracy, high-detailed data ensures precise planning and decision making. And the flexibility, I think, is always given by complex 3D scanning workflows. This video showcases a point cloud autonomously captured in a chemical plant, highlighting the exceptional level of details achieved with the BLK ARC. The scans reveals features ranging from concrete pillars and tanks, to ceiling installations, building foundations, and much more. Scanning is equally effective indoors and outdoors. However, it's worth to mention that rain is not a friend of scanning, as it can influence the point cloud. Towards the end of the video right here now, you have seen the example to transform the point cloud into mesh and even how to create models out of it. The key applications, for example, in the chemical context is infrastructure analysis, like detailed scans of the plant, layout, and components, helping engineers assess structural and operational integrity, and planning and retrofitting. I think I mentioned already before, maintenance planning. It's a very common daily path and task for all of you guys. The point cloud provides a visual measurable data set to identify the wear and tear of potential hazards in the ceiling installation tanks or foundations, for example. And finally, safety compliance. I think that is as mentioned before. And you can do even more with the point cloud. The point cloud plays a crucial role in enabling real time renderings by providing highly detailed and accurate spatial data that can be processed and visualized in real time. Facility walkthroughs will be done on a new level. Facility managers or planners can explore the digital twin of the site in real time. Inspection details, and planning layouts interactively. So for example, in NVIDIA Omniverse, you can bring the point cloud back to life. And collision detection, you can simulate the movement of equipment or personnel in real time. Augmented reality, virtual reality, makes suddenly sense because you have a real-time representation of the environment. And this provides you an immersive experience, by serving as a foundation for the virtual reality. Laser scanning significantly enhances building information modeling, so-called BIM, by providing precise and detailed and comprehensive spatial data. Some examples for this BIM, scan to BIM, how we always call it, is construction verifications. So scans can be used throughout the construction process to compare ongoing work against the BIM model, ensuring adherence to design specifications. This helps to detect deviations early and prevent costly delays and repairs or modifications. And hence, facility management. And the digital twin creation, the laser scans enable the creation of a digital twin and the dynamic, up-to-date virtual replica of the building or the site, which is central to BIM workflows. And facility managers can use the data for maintenance planning, asset tracking, and lifecycle management. Improvement in design and renovation planning, so existing conditions are scanned. So the scans can be integrated into the BIM to analyze existing infrastructure before designing modifications or additions. Collision detections, high-detailed scans allow designers to identify potential clashes between new designs and existing structures. For example, pipes and walls, they don't like each other a lot. So I think it's common in construction industry already that BIM and scan are each other. And again, imagine you can do this on fully automated. Another use case, what we really get a fairly large customer group in the meantime, is anything about power distribution. I think autonomy is not only an advantage of using a robot. Sometimes remote control is equally valuable. This approach allows the robot to be sent into areas that pose potential risks to humans, such as contaminated zones within a nuclear plant, I think very obvious. But also, a fully operational high voltage substation could be a hazard for a person. Laser scanning of substation is an invaluable tool for capturing detailed, accurate, and comprehensive spatial data. It supports a wide range of applications in design maintenance operational workflows. SPOT, the most dynamic SPOT, SPOT ensures a remote and safe data collection. It captures data without requiring personnel to enter high risk areas, such as high voltage environments zones within the substation. As mostly and most important, it reduces the downtime as scanning can be performed by the substation remains fully operational. Use cases similar again, is for as-built documentation, modeling, design and planning, and many more. One of my favorite elements when I do my POCs is always underground. I'm not an underground surveyor. So for me, it's an adventure if you go 1,000, 1,500 meters below the surface and scan there. Autonomous laser scanning transforms, tunneling and mining operations by providing efficient, safe, and highly-detailed data collection in challenging environments. Safety in hazardous environments is key. That is, when you go into a mine, there is always one rule, zero accidents. That's all about zero accidents. And with SPOT, we have the power to continue increasing or ensuring this low-risk for our personnel. So the data is collected even in unstable low visibility or in areas at risk for collapse without putting personnel into risk. I think a huge use case. And again, efficiency in coverage by operating almost continuously while operating in these environments is another key benefit for robot. You need to, and Brian already mentioned it before, classification. I think here it's a short example of how classification look like. And it's about adding easily, intelligence to the point cloud. In our Cyclone 3D software, you can see, as in the video, by simply clicking one button, you can select the type of classification you want to run. In this, I want to use people. And by a simple mouse click, these peoples get filtered out. You see, those are identified. These are all moving objects. And like in the CAD layer, you can just reduce or release this feature if you are not interested. So classification of point clouds is essential for extracting meaningful insights and organize the data for specific applications. Classification assigns A-defined classes like wall, floor, pipe, vegetation to each point in the point cloud, making the data interpretable and actionable. It reduces manual effort of cleaning and escalate the workflow, especially in complex environments like factories or construction sites. One of many benefits is to remove unwanted objects easily. I think if you want to work with a BIM, you don't want to see plants or vegetation. And this can be reduced very straight and forward. With that, thank you very much for giving me the time and the opportunity to explain what we do. And I would like to hand over to Brian. - Yeah, thank you so much, Johannes and Julita. That was great. A lot of really specific detail, and all the application stuff is really powerful for downstream work. These aren't just point clouds. It's semantically-labeled information, which is critical. And the three of us could talk all day about this technology. But I think it's better for you to hear it directly from our customer, Marc Hemling at ST engineering. So we'll queue up a video on that. [VIDEO PLAYBACK] - When you want to ask what SPOT's doing for us, it's gathering data that we didn't have before. And it allows us to build a database filled with records of critical equipment, statistics that we can go and eventually use to diagnose or prevent machine failures, to help keep uptime, and even safety of the people that work around the machines. Being in a facility that is 95 years old, we do have engineering drawings of the facility. But they don't really capture everything. We do our best to maintain them, but there are components and things that may have been left from 60 years ago that are still here and not captured. And it makes it awfully tricky as we bring in new equipment and technology to understand what exactly is in the area we're trying to move this equipment to. Not only is it beneficial for your current projects, but as we go and we make changes and we are constantly adding new equipment in here, keeping up with that is very hard. And then not only keeping up with it, but ensuring that it's done accurately, that's where the scanning comes in, is it allows us to take a snapshot in time and something that you could walk through and see the current status of it and then see the changes that we make. But for me, a digital twin, it's a 3D representation of the current state of your facility. And SPOT not only does that with the inspections, but now with the added capability of the Leica scanner, we can now create a 3D model of our facility and also inspect it to keep our machines operating properly. The real value of the combination between SPOT and BLK ARC is the time that it saves because the mobile scanning ability of the ARC, but then the automated missions of SPOT allow for the saving of time that no one here has. I think it's something that any company that is serious about having a digital twin of their factory, they will consider it, especially if they have SPOT already. Something that relatively easy to do, they can even use existing machines and convert it. Anyone that has felt the pain of trying to scan and wants to keep a digital twin will be looking at this technology. [MUSIC PLAYING] [END VIDEO PLAYBACK] - So I got to visit that site. And it was really cool to be there and watch one of our first customers expand from the industrial inspection package with the sensors to the BLK ARC and start to add that 3D environment scanner to the repertoire and to watch the Leica BLK ARC robot and the industrial inspection package robot have missions scheduled simultaneously to see them both pop up off of the dock and go their separate ways to be combining that information in Orbit was really powerful. So again, like I said, I think this is just the beginning of what you can do with digital twins. As somebody who's been working in the reality capture and digital twins space my entire career, I wouldn't have thought of it as being uniquely enabled by this new class of agile mobile robots, but here we are. It's really something that's been powerful for our customers and provides a really kind of exciting future for us. So we're going to continue to work on these features, continue to combine all of these data streams for that comprehensive, holistic, up-to-date, real-time vision of your factory operations and the facility, the physical facility itself. And this is all part, I think, of the value that we deliver with the SPOT robot, with Orbit when it comes to better operations, more savings and operations, safer operations, removing people from situations where they don't have to be. And I think we're really, really proud of all that work. So again, thank you to everyone who took the time out of their busy days to attend this and learn more and speak with us. And now I'm going to turn it back over to Michaela to moderate some Q&A. - Yeah, thank you everyone. - I just want to give a little teaser reminder, if you don't mind. We are having a laser scanning-specific seminar or the autonomous laser scanning, specific focused weapon on the November 11 next year, where we will deep dive more into the technology, into use cases and give some more insights on that one. Please feel free. We are happy if you join us there. The QR code gets you to register on the page. So maybe we see us all again in the new year. - Much more to come. Well, we have a lot of great questions. Let's start with-- we have a few questions along the lines of different applications for SPOT and using other payloads, alongside the BLK ARC. Maybe Brian, can you speak to what we'd recommend for somebody who might currently have SPOT with other sensing payloads or is interested in doing some different things with their robot? - I mean, certainly at the end of the day, payloads are configurable. You can swap them out. But that requires somebody to be there at site. And I think the beauty of the system with Orbit, with the automatic scheduling of missions from these docks that can be done remotely and autonomously, the value would be to set up dedicated robots for each of these things. That also allows you to run these things, as I mentioned, with the customer example there. These things can then be run in parallel to maximize your data operations. So yeah, we do recommend that you maintain these as a separate robot, separate payloads. And then that way, they can be autonomously deployed without field intervention as needed. - And while we're talking about payloads, can you walk us through the difference between use cases for the EAP-2 versus the BLK ARC? Does the BLK ARC replace the EAP-2? - Yes. So some of the savvier among you noticed that there are technically two LiDAR payloads on the robot in that scenario. And the deal is that the LiDAR payload that is offered by Boston Dynamics, which is part of the EAP-2 or enhanced autonomy payload package, is for navigational purposes only. The data that comes out of that is extremely coarse because that's what's computationally efficient for the robot's navigation and autonomous localization. And it's also not registered in a globally accurate way, which makes it not very good for surveying use cases, even though it works quite well and reliably for autonomous navigation. So those things are kept separate. The BLK ARC data is of a much higher quality. And that's why we see that as needed for a dedicated payload. And there are ways to mix and match too. Leica maintains a separate solution for our non-industrial inspection and Orbit customers, where they use the Boston Dynamics SDK to converge those two things together, the data capture and the autonomous navigation. But in the package that we're describing today, those are two separate things. And they are each doing unique and valuable jobs on the back of the robot. - Thank you. We have a few questions about data processing. First up, let's see. What are the output files and post post-processing requirements? Julita, maybe this is a good one for you. - Yeah. So earlier I mentioned that the direct output file from the BLK ARC file is the B2G, which is our native file format. However, after processing the data, the point clouds can be exported into more open files, such as E57, BTS, and RCP, which many, many different software can actually digest. - Worth to mention. I think, that you even can convert on SPOT I/O B2G into an E57. - Johannes, can you talk us through how the data is uploaded? We had someone asking about why it's not uploaded continuously while the robot is walking the facility. - The BLK ARC creates a live point cloud. And this live point cloud can be streamed. And we have customers who utilize this data stream live to control or create environments in real time. However, if you want to have a full dense point cloud, this data stream requires a lot of bandwidth. And this is very often a challenge. So if you want to stream the full 420,000 point cloud, typically, our users do it in post-processing. But theoretically, yes. And we have customers who use the live stream as well. - And I might add some color to that by saying that when we design our systems, they're designed to be able to perform autonomously without a network connection. Because it's really difficult for our customers to have good wireless connectivity throughout all of the places that they're trying to inspect. So it's really important that the robot can drop comms and continue doing its work. So we always default to a way where when the robots out walking around, it does not need to be connected. And it can keep that data stored locally. And then there's inevitably there's some recharge time on the dock. We advise customers to set those dock locations at places that do have good connectivity wirelessly, or even use the ethernet passthrough for wired connectivity. And then that, that really gets you the best of both worlds. But as Johannes says, it is possible. And it's something we're thinking about in development. - For someone who's interested in generating a 2D floor plan, can you recommend the best software for converting their LiDAR scan results into CAD? - I think there is a various different options. Typically, you do a slice of our point cloud that you see, that's just the walls and the ground. And then you can utilize software like the cyclone 3DR, who does then automatic, half-automatic, the creation of floor plans. But there are also other software packages, like a CAD package, like BricsCAD or AutoCAD, who then can utilize the point cloud and create a floor plan out of that one. And maybe just worth to mention, we have mentioned Cyclone 3DR. But we also have something like called Cloud Works. Cloud works are add-ons for a wide variety of different software packages. So we have Cloud Works for AutoCAD. So that's an add on to AutoCAD, where we can have our point cloud tools within AutoCAD or Navisworks or whatever. So we have a wide variety of different packages are available there as well. So I would go for a Cyclone 3DR or BricsCAD to create a 2D floor. - And I also can-- we've mentioned it multiple times. But the auto-classification models in 3D are really amazing. So for example, when we do indoor data capture, we can run the indoor class model. And you get separation of furnishings from any system. Then you have your data semantically separated so that you can put them on the right CAD layers. You can remove people and other mobile objects that might be noise that you don't care about. It just makes it a lot easier to slice the model and know what's what and have it layered in a way that makes sense for a CAD workflow. - Thanks. Obviously, this presentation was all about our partnership and integration with Leica. But we had a couple people asking if there are other 3D scanning solutions that can be used with SPOT. - Yeah, the short answer that I give everyone that may or may not be helpful is anything with an API can be connected to SPOT if it's necessary for your business. But I'll hit on this again because I mentioned it briefly in the beginning. Not all laser scanners, and in fact, very few are designed for mobile transport on these types of walking robots. So first and foremost, not even thinking about the software features and the other reasons for working with Leica, this is hardware that's designed for the back of this robot. Sometimes when I talk about scanning with robots, I try to get people to see it, not as robot plus scanner, two things combined together, but as an entire new class of hardware, where we're kind of taking the best of both worlds and trying to make something new that's greater than the sum of its parts. I have worked with other scanners that-- there are a lot of scanners out there. They all do really great things. And I've used a lot of them. And this is nothing against them, but they have to be designed to withstand certain types of internal forces for sensitive objects that allow scanning to be done. So the fact that it's mobile, that you get the right resolution for the right time, that's always kind of a trade-off, precision, resolution, time, coverage, and the fact that it's specifically designed to be compatible with a walking machine, essentially really, are what made it the right fit for bringing into Autowalk and Orbit. - And for just a little example, why design for robotics. The BLK ARC does not have a battery. It is powered by SPOT. So if SPOT is charged, BLK ARC does not need replace of the battery. So all these continuous operations, I love the new feature that multiple docking stations are now supported. And the arc doesn't care because it gets powered by SPOT. Then secondly, it's the weight, 690 grams of the BLK ARC. So glass of beer may be maximum. And this is designed, not to make it small. It's designed that the battery consumption, the power consumption of the SPOT is minimum influenced when carrying the BLK ARC. And it's smaller designed that other payloads are fit onto it. And so, BLK ARC, you see it, we have a front mount, we have a core mount, we have a central mount. So we do provide different mounting options even when the arm is, for example, installed. But you still can add other things. So that's the design. The overall design of the BLK ARC is to make it small and compact, not to influence anything on the robot. And I think that's the key element of a mobile scanner or scanner for a robot. - And I think it's worth mentioning, one of the very first things we tried with SPOT, even before SPOT was a commercial product, was laser scanning integration. It is one of the original applications that people have been asking for. That means we have a lot of experience with a lot of hardware, a lot of workflows, a lot of partnerships. And so this is a very educated decision that this is the best offering to bring to bear for the industrial inspection package. - Thank you. I think we have one time for one more quick question. How accurate is the mapping with the BLK ARC? - That's always the classical. There are two questions always, how fast it is and how accurate it is. So accuracy, it's a SLAM-based scanner. So accuracy depends on a little bit on the environment. But in a SLAM-friendly environment, which is very typically for construction sites, for industrial facilities, because there are a lot of features, a lot of geometry. You rate this one as SLAM-friendly. So typically the absolute accuracy of an entire facility is something between 15 to 25 millimeter. Sorry. You hear it on my strong Texas accent. I convert that one to the inch or a little bit less. Relative accuracy, and this is always-- I see the entire global accuracy. But very often you are interested in the relative accuracy or the width of the door or size of this room, and that is typically in the range of a millimeter or 1/16 of an inch. Time saving is enormous. Julita said something like 50% Typically, I hear it from our customers. It's even 5 to 7 times faster. So the more complex, more machinery is in the industrial hall, the more times we save. Because terrestrial laser scanning requires them many, many setups. If it's one large open room, the advantage in time savings are a little bit less. Because then the transistor laser scanner sees from one position a lot. But more installation, more rooms, more blockages for the laser, the higher is the time savings. And again, I said 5 to 7 times. And here, I've not taken the advantage of autonomous scanning with SPOT. Because one SPOT has learned the environment, then every other scan at the same location is just the press of a button of the operator in Orbit, which is absolutely fantastic. I'm really amazed by that. - Awesome. Well, with that, thank you so much to our presenters. Johannes, Julita, Brian. We'll be sending out the recording after we wrap up today and following up with anyone who asked a question that we didn't get to today. So thanks, everyone, for attending. And we'll see you next time. - Thank you very much. - Thank you.